[clang] [llvm] [AMDGPU] Add support for `v_cvt_f16_bf8` on gfx1250 (PR #146305)

[Shilei Tian via llvm-commits]

================
@@ -2,51 +2,41 @@
 # RUN: llvm-mc -triple=amdgcn -mcpu=gfx1250 -mattr=+real-true16 -disassemble -show-encoding < %s | FileCheck -check-prefixes=GFX1250,GFX1250-REAL16 %s
 # RUN: llvm-mc -triple=amdgcn -mcpu=gfx1250 -mattr=-real-true16 -disassemble -show-encoding < %s | FileCheck -check-prefixes=GFX1250,GFX1250-FAKE16 %s
 
-0x05,0x00,0xf2,0xd5,0xc1,0x00,0x00,0x00
-# GFX1250: v_cvt_f32_bf16_e64 v5, -1               ; encoding: [0x05,0x00,0xf2,0xd5,0xc1,0x00,0x00,0x00]
----------------
shiltian wrote:

https://github.com/llvm/llvm-project/pull/147521

https://github.com/llvm/llvm-project/pull/146305